Odour Recognition and Segmentation by
Coupled Olfactory Bulb and Cortical Networks
Zhaoping Li
Gatsby Computational Neuroscience Unit
University College London
John Hertz
Nordita, Denmark
Neurocomputing 26-27, 789-794 1999.
This work is presented at CNS conference 1998.
Abstract
We present a model of a coupled system of the olfactory bulb and
cortex. Odor inputs to the epithelium are transformed to oscillatory bulbar activities.
The cortex recognizes the odor by resonating to the bulbar oscillating pattern when the
amplitude and phase patterns from the bulb match an odor memory stored in the
intracortical synapses. We assume a cortical structure which transforms the odor
information in the oscillatory pattern to a slow DC feedback signal to the bulb. This
feedback suppresses the bulbar response to the pre-existing odor, allowingsubsequent odor
objects to be segmented out for recognition.
Download [ps.gz] [pdf]